Hypoxia and Environmental Epigenetics

Introduction

Epigenetic changes are heritable alterations in chromosomes that, unlike mutations, do not involve changes to the DNA sequence, but rather represent structural modifications to the nucleic acid:protein complex that makes up chromosomes. Transmission of epigenetic marks during cell replication allows descendent cells to maintain the same patterns of expression, and thus differentiation status, as their progenitors, thereby permitting clonal descent of specialization. For example, a dividing epithelial cell will give rise to a second epithelial cell, despite the fact that the nascent cell has all of the genetic information required to become any type of cell in the body. The mechanisms underlying this form of ‘cellular memory’ are not fully understood but a number of epigenetic ‘marks’ have been identified, including methylation of cytosine bases and modification of histone proteins. The effects of these marks are to alter the accessibility of genes to the transcriptional machinery and, by doing so, repress or activate expression of the gene. At the organism level, these changes will drive, and then maintain, cell differentiation, as well as modulate transcriptional responses to changes in the environment. Recent data show that epigenetic changes that occur during embryogenesis or early in childhood can modify gene expression, and thus the organism's phenotype, throughout life. Such mechanisms would be valuable tools for acclimatization or developmental adaptation, allowing an organism to reversibly optimize their, or their offspring's, ‘future’ gene expression, for specific environments.

A number of hypoxia-induced genes are epigenetically regulated. While much research is focused on identifying genetic variants that affect the physiological and pathophysiological response to hypoxia in humans (reviewed in MacInnis et al., 2010), the possibility that epigenetic differences contribute to inter-individual variation in hypoxia tolerance needs to be considered as well. Such differences could manifest acutely in response to exposure as part of the acclimatization response, chronically in response to long-term exposure, through an in utero response to maternal hypoxia stress, or potentially trans-generationally through germ-line modifications. As epigenetic changes are reversible, this programming could be transient, with the system reverting to pre-exposure levels upon descent. Alternatively, the changes could be retained upon descent, facilitating acclimatization during subsequent exposures.

Epigenetic Markers: DNA Methylation and Histone Modifications

In humans, the epigenetic marks that maintain genes in quiescent states include DNA methylation, histone protein modifications and variants, and functional noncoding RNAs (ncRNAs) such as long ncRNAs and microRNAs (miRNA). In vertebrates, DNA methylation primarily occurs at CpG dinucleotides (i.e., on a cytosine that is followed by a guanine with a phosphodiester bond between them). The CpG is palindromic (i.e., the alternate strand is also CpG, but in the other direction), which allows for stable inheritance of DNA methylation during replication by a maintenance DNA methyltransferase (Fig. 1). In addition to being maintained, methylation patterns can be altered by addition (by DNA methyltransferases), or removal, of methyl groups. The latter can occur passively (by blocking of maintenance methylation) or actively through the activity of the ten-eleven translocation (TET) family of enzymes and the recently discovered 5-hydroxymethylcytosine (Kohli and Zhang, 2013). Approximately 70% of mammalian CpGs contain a methylated cytosine; however, due to the high mutation rate of methyl-cytosines, the CpG dinucleotide is substantially under-represented in the mammalian genome. CpGs are not evenly distributed throughout the genome but rather are clustered in ‘islands,’ usually near the promoter regions of genes. These islands are generally unmethylated in most tissues. However, methylation of these regions correlates with reduced expression of the adjacent gene, while the impact of DNA methylation beyond island promoters is less well understood (Weber et al., 2007).

OPEN IN VIEWER

FIG. 1.

Epigenetic mechanisms that have been implicated in the molecular response to hypoxia. Removal of acetyl groups (A) from histones and/or addition of methyl groups (M) to cytosine bases (in CpG dinucleotides, white circles) tend to condense chromatin, thereby altering the accessibility of genes to the transcriptional machinery and impeding transcription. While histone acetylation has been studied in response to hypoxia, there are a variety of other histone modifications including methylation, ubiquitination, and phosphorylation that potentially could be involved as well. DNMT, DNA methyltransferase; HAT, histone acetylase; HDAC, histone deacetylatase. Thick black line is DNA, spheres are histone proteins (H2, H3, and H4 in nucleosome, H1 in linker DNA).

Chromosomes are a complex of DNA packaged into a linear array of nucleosomes, which are short (about 150 bases) stretches of DNA wrapped around a protein octamer comprised of two each of the four core histones (H2A, H2B, H3, H4). The DNA linking the nucleosomes is associated with a fifth histone (H1), and the resulting DNA protein complex is referred to as chromatin. Modifications of the histones or the substitution of ‘standard’ histones with histone variants (e.g., H3 variant CENPA) can alter the accessibility of the DNA to the transcription factors and enzymes required to read the genes, thereby modifying expression (Fig. 1). While the most commonly studied histone modifications are acetylation and methylation, there are numerous others that may contribute to epigenetic regulation (e.g., phosphorylation, ubiquitination; reviewed in Bannister and Kouzarides, 2011).

Recent developments in high-throughput transcriptomic analyses have demonstrated that a large percent of RNA molecules transcribed from the mammalian genome are not translated (reviewed in Costa, 2010). Small non-coding miRNAs act primarily to negatively regulate the translation of messenger RNAs by binding, thus marking the mRNA for degradation or repression. While miRNAs have been shown to be involved in the molecular response to hypoxia (reviewed in Jacobs et al., 2013), whether long ncRNA-mediated epigenetic control contributes to hypoxia adaptation has yet to be investigated. A well-studied example of RNA-based silencing is X-chromosome inactivation in which dosage balance between XX females and XY males is achieved through the action of the long non-coding RNA XIST, which, early in development, marks for inactivation one of the two X chromosomes in females (reviewed in Chow et al., 2005). A number of X-linked genes have been shown to be regulated by HIF-1 (Mole et al., 2009) or implicated in susceptibility to altitude illness (MacInnis et al., 2010), and a hypothetical X-linked gene has been postulated to play a role in women's greater resistance to hypoxia-related mortality (Mage and Donner, 2006). If variants in X-linked genes contribute to hypoxia tolerance, the phenotype may be influenced by the inactivation process, as males, who are hemizygous (i.e., carry only a single copy of the gene) will express the variant in all cells, while heterozygous females will be mosaics with the variant on the active chromosome in only a subpopulation of their cells. Furthermore, an estimated 15% of X-linked genes escape inactivation and therefore may have higher expression in females than males.

Changes in DNA methylation patterns or histone modifications alter the expression patterns of genes and, despite the shifts in the functionality of the region, the actual sequence of the DNA remains unchanged. Epigenetic differences between two phenotypically different cells, individuals, or populations are not distinguished by current genotyping or DNA sequencing methods, which complicates the search for epigenetic contributors to complex traits such as altitude acclimatization. Even if a gene did contribute to the variability in hypoxia tolerance, the differences between altitude-sensitive and altitude-resistant people could be epigenetic rather than due to changes in the sequence of the gene. Alternatively, an epigenetic change could mimic the effect of a sequence variant (e.g., phenotypically, an epigenetic change that reduced expression would be similar to a sequence variant that reduced expression).

Superimposed on this complexity is the possibility of sequence variants that alter methylation patterns by generating or eliminating CpGs. Current data suggest that epigenetic marks occasionally can be passed between generations, providing a molecular mechanism by which the environment can impact the expression of genes years to decades later, allowing offspring to be “pre-programmed” to deal with stresses experienced by the parents (reviewed in Youngson and Whitelaw, 2008).

Transgenerational Transmission of Epigenetic Information

In rats, maternal stress will affect pup behavior, even if the pups are fostered to nonstressed mothers at birth. This pre-partum programming involves alterations in DNA methylation and acetylation of the chromatin flanking the glucocorticoid receptor gene, which results in decreased glucocorticoid receptor gene activity in the hippocampus (Francis et al., 1999). Similar changes at a subset of CpG dinucleotides in one of the alternative promoters of the glucocorticoid receptor are associated with childhood abuse in humans (McGowan et al., 2009). Maternal diet can also affect epigenetic regulation in the offspring. Coat color in the Agouti viable yellow (A^vy) mouse is influenced by the insertion of a retroelement (a fragment of repetitive DNA) that provides an alternative promoter to the Agouti gene. Methylation status of CpG dinucleotides in this promoter affects its function, resulting in a range of coat colors from pale yellow to the wild-type brown (Morgan et al., 1999; reviewed in Morgan and Whitelaw, 2008) despite the animals having identical DNA sequences. The color of A^vy offspring usually correlates with that of the conceiving mother, even if the pup is transferred to a different colored female as an embryo; however, altering the maternal diet can change the color or the offspring. This color change ‘breeds true’ in the sense that the next generation will resemble the first generation offspring and not the ‘grandmother’ (Martin et al., 2008).

The sensitivity of the A^vy mouse coat color to exogenous insult is sufficient for the animal to be used as a screen for environmental toxins that act by altering epigenetic patterns (Bernal and Jirtle, 2010). Maternal dietary supplementation with genistein, the soy-derived phytoestrogen, has been shown to shift the coat color towards agouti (Dolinoy et al., 2006) while mothers fed bisphenol A have lighter colored offspring. The latter effect can be counteracted by dietary supplementation (Dolinoy et al., 2007), demonstrating the complexity of the epigenome:environmental interactions.

Responses to changes in environment can also be transmitted beyond the next generation. Rats fed protein-restricted diets during pregnancy exhibited elevated blood pressure and vasculature endothelial dysfunction, as did their offspring. Both male and female grand-offspring (through the maternal line) showed similar characteristics, despite the fact that their mothers (from the affected FI generation) were fed a normal diet (Torrens et al., 2008). The nutritional phenotype was accompanied by hypomethylation of gene promoters, suggesting an epigenetic effect (Lillycrop et al., 2008) and reduced transcription (Lillycrop et al., 2010), notably in the Gene Ontology ‘Biological Processes’ category “oxidative stress.”

Relevant to environmental hypoxia research, the endothelial dysfunction was due to deviations in the nitric oxide (NO)/cGMP pathway, which is believed to play a role in acclimatization and adaptation to altitude (reviewed in Janocha et al., 2011; Beall et al., 2012). The promoter region of the gene encoding eNOS is heavily methylated in nonendothelial cells, while in endothelial cells, the same bases are unmethylated and the associated histones modified to facilitate transcription (reviewed in Illi et al., 2009; Yan et al., 2010; Illi et al., 2011). Levels of methylation in the hypoxia response element in the NOS3 promoter were associated with differences in expression of the gene in umbilical blood vessels between control and growth-restricted fetuses.

One of the best known examples of epigenetic adaptation in humans comes from studies of the children and grandchildren of women who survived the ‘Hongerwinter’—the 1944–1945 Dutch winter famine. The mothers who gave birth in Amsterdam between January and December 1945 (i.e., who conceived and carried during the famine) weighed less than women who conceived and carried in the flanking years (the ‘control’ women) and the children who were exposed to the famine in utero were smaller and lighter at birth than were the children of the control women and, as adults (age 50), had lower glucose tolerance (Painter et al., 2008). In the second generation, children whose mothers were exposed to famine in utero were smaller and less ‘lean’ (i.e., higher ponderal index) and had, by broad definition, “poorer health.” These second generation effects were not transmitted through the paternal line; however, children of prenatally exposed fathers had a higher BMI than those of the controls (Veenendaal et al. 2013). These data suggest that environmental changes can elicit metabolic changes in subsequent generations but the resulting phenotypes are affected by the parent through which the signal is passed. The mechanisms by which this information is transmitted, including the extent to which it can be maintained through gametogenesis. have yet to be fully worked out but there is evidence that epigenetic factors may be involved. Six decades after prenatal exposure to deprivation during the ‘Hongerwinter,’ both men and women had lower levels of methylation of the imprinted IGF2 (insulin-like growth factor-2) gene than their same-sex siblings (Heijmans et al., 2008). There is also evidence that grandparental behavior can be transmitted in a sex-specific manner through two generations. Pembrey et al. reported that grand-paternal and grand-maternal nutrition status effected mortality risk ratios in grandsons and granddaughters, respectively (Pembrey et al., 2006) and a similar observation was made for longevity in the same lineages (Kaati et al., 2007).

The ability of an organism to ‘reprogram their offspring’ in response to changes in the environment is a paradigm-shifting revelation with respect to understanding acclimatization and adaptation. Much of the current data in mammals is related to changes in nutrition; however, similar mechanisms may have facilitated human migration to high-altitude environments. Transgenerational uterine reprograming would greatly confound studies that use migration and fostering to parse out the genetic contributions to complex traits. Even children conceived and born at sea level to high-altitude mothers may be manifesting effects of the hypoxic stress experienced by their grandparents as much as 40 years earlier.

Epigenetics and Hypoxia Detection

The carotid bodies monitor oxygen levels in the blood; even minor departures from normoxemia will trigger increased rates of sensory nerve firing. Perinatal exposure to intermittent hypoxia alters the sensitivity of these sensors (reviewed in Prabhakar, 2013). This response may be mediated epigenetically. The carotid bodies and adrenal medulla of adult rats who had been exposed to intermittent hypoxia for the first 10 days post-partum showed higher levels of the DNA methyltransferase enzymes Dnmt1 and Dnmt3b as well as the encoding mRNAs (Nanduri et al., 2012). Neonatal exposure to intermittent hypoxia altered adult carotid chemoreceptor sensitivity but the effect could be abrogated by the potent DNA methylation inhibitor decitabine (which blocks DNA methyltransferase, Fig. 1). Intermittent hypoxia was also associated with methylation of a CpG near the transcription start site of SOD2 (the gene encoding antioxidant superoxide dismutase 2, SOD2) in adrenal tissue and with lower expression of SOD2 and lower levels of the SOD2 enzyme. Again, these effects could be blocked with decitabine.

Epigenetics and Gene Expression in Hypoxia

The transcriptional response to hypoxia is initiated by the binding of HIF-1 (hypoxia-inducible factor 1) to a hypoxia response element (HRE) site. HREs are recognition sequences of DNA proximal to the gene under hypoxia regulation (reviewed in Semenza, 2012).

The sequence of HREs varies but all contain the core sequence CGTG. Regulatory regions of hypoxia induced genes contain multiple core sequences, sometimes arranged in tandem, as well as other transcription regulatory sites. Wenger et al. lists the HRE sequence for over 70 human genes (Wenger et al., 2005), a number of which are shown in Table 1. In all cases, the CGTG core sequence is flanked by CG-rich regions. As discussed earlier, mammalian methylation occurs only at CpG dinucleotides, one of which occurs in the HRE sequence, suggesting that the HIF-1 regulated gene expression could be sensitive to cytosine methylation and therefore epigenetic regulation. In addition to DNA methylation modulation of HIF-1 target genes, histone modification may play a role in the molecular response to hypoxia. HIF-1α is complexed with several histone acetyltransferases that may play a role in activating HIF-1 regulated genes (reviewed in Semenza, 2012). Once activated, HIF-1 upregulates several histone demethylase enzymes, which act in conjunction with the HIF dimer to activate genes (Watson et al., 2010).

EGLN 1 encodes PHD 2, the principal regulator of HIF-1a. Recent studies have implicated the gene EGLN1 in human evolutionary adaptation to altitude (Simonson et al., 2010). EGLN1 has an upstream HRE (GGTGTACGTGCAGAGCGC) and the gene is sensitive to upstream cytosine methylation, as is VHL (another core component of the molecular hypoxia response). EGLN1 is also sensitive to histone modification secondary to environmentally toxicity, for example, metal (nickel) increases the frequency of H3K4me3 (trimethylated histone H3) in the promoter regions of EGLN1, which is associated with elevated expression of the gene (Chervona et al., 2012).

EPO, which encodes the glycoprotein hormone erythropoietin, is likely the best characterized of the hypoxia-induced genes. The HRE in EPO is downstream of the gene and not associated with a CpG island; however, there is a CpG island in the upstream promoter region and, during transcription, the gene folds back on itself, allowing the HIF-1 transcriptional complex to interact with both the 3′ HRE and the 5′ promoter regions simultaneously. Methylation of either regulatory site will inhibit transcription of EPO (Rossler et al., 2004; Yin and Blanchard, 2000) and methylation of the upstream regions may be involved in establishing tissue specificity of the gene. Hypoxia altered the methylation patterns of the EPO promoter in the Hep3B cell lines examined by Yin et al. (2000) suggesting that hypoxia-induced epigenetic changes could alter expression of EPO. In cell lines with unmethylated EPO promoters, EPO gene expression was significantly increased in response to hypoxia. In lines with a methylated promoter, the gene was nonresponsive but could be rendered inducible by treatment with a DNA methyltransferase inhibitor (Steinmann et al., 2011). In addition to the regulatory effects of cytosine methylation on the gene, hypoxia-induced EPO expression is also known to be modulated by acetylation of histones H4 in the promoter region of the gene (Wang et al., 2008).

In addition to epigenetic regulation of EPO, other genes in the erythropoietic pathway may also be under epigenetic control. Erythropoietin increases RBC proliferation by inhibiting apoptosis in the erythroid cell progenitors (Jelkmann, 2004), thereby allowing more cells to mature. Histone deacetylases can inhibit this process (Yamamura et al., 2006).

The HIF transcription complex includes a number of co-activators (e.g., p300, CBP) that facilitate reading of the target genes by ‘relaxing’ the local chromatin thereby making the DNA more accessible to the enzyme (reviewed in Perez-Perri et al., 2011). CBP and p300 are histone acetyltransferases that can increase or decrease hypoxia-induced expression from a number of key hypoxia response genes (including EPO, see Perez-Perri et al., 2011).

Hypoxia-induced changes in gene expression likely extend well beyond the HIF-induction pathway. Hypoxia can cause genome wide alterations in histone acetylation and CpG methylation (Perez-Perri et al., 2011), potentially resulting in global modification of transcription levels (Johnson et al., 2008).

Epigenetics and Altitude-Related Pathologies

While there has been some speculation that epigenetics could play a role in the susceptibility or etiology of altitude sickness (e.g., Luo et al. 2012; MacInnis et al., 2010), no studies have tested this postulate (although epigenetic studies are planned for Operation Everest II; see: http://www.xtreme-everest.co.uk/article.php?article=847). Hypoxia induces Brg1 and Brm in cultured human endothelial cells that in turn epigenetically upregulate expression of cell adhesion molecules (CAM; Chen et al., 2013). Brg1 and Brm encode chromatin-remodeling complexes that are known to regulate the transcriptional responses to hypoxia (Sena et al., 2013). As increased CAM in the pulmonary endothelial airways favors leukocyte adhesion and therefore an inflammatory response, these data demonstrate an epigenetic role in the development of hypoxic pulmonary hypertension. Suppression of endothelial Brg1 and Brm genes in mice reduced right ventricular hypertrophy in mice exposed to simulated altitude (P_B ∼405 mmHg, 4 weeks; (Chen et al., 2013), suggesting that the treatment reduced hypoxic pulmonary hypertension. Humans with an exaggerated pulmonary-tensive response to hypoxia are more susceptible to high altitude pulmonary edema (HAPE; Scherrer et al., 2006), but whether the same genes and epigenetic pathways contribute to development of (or susceptibility to) the condition has yet to be tested.

Demonstrating a connection between epigenetic changes in the pulmonary epithelium and HAPE in humans would be challenging unless the changes were reflected in peripheral (i.e., readily accessible) cells.

Epigenetics and Twin Studies of Hypoxia

The use of twin studies to unravel the environmental and genetic contributions to complex traits is based on the premise that monozygotic (‘identical’) twins are genetically identical, and therefore differences between members of a twin pair must be due to environment. While this is true with respect to the sequence of the DNA (barring mutations), there can be epigenetic differences that contribute to discordance. Petronis et al. reported less variation in patterns of DNA methylation in the D2 dopamine receptor gene between twin-pairs concordant for schizophrenia than between those discordant for the condition (Petronis et al., 2003). Similarly, twin pairs discordant for systemic lupus erythematosus showed differences in the methylation patterns of genes thought to contribute to the condition (Bernal and Jirtle, 2010).

Epigenetics may also contribute to a greater overall discordancy between female twin-pairs and male pairs. The canonical example of epigenetics in mammals is the process of X chromosome inactivation, through which one of the pair of X chromosomes in XX females is silenced, thereby achieving dosage equality for X-linked genes with XY males (reviewed in Chang et al., 2006). The initial inactivation event occurs at around the eight-cell stage and is random with respect to the parental origin of the X chromosome (i.e., maternal or paternal); however, once the ‘choice’; is made, all progeny cells will maintain inactivation of the same X chromosome. This results in identical female twins-pairs potentially differing in which X chromosome (the maternal or the paternal) is active in any given cell, and if heterozygous at an X-linked locus, which allele is expressed. For this reason, female monozygotic (MZ) twin pairs would be expected to be more variable for any characteristic that is influenced by genes on the X-chromosome than would male pairs.

Twins also provide evidence that epigenetic patterns can be inherited, as there is a greater correlation between epigenetically-regulated gene expression in MZ twins compared to dizygotic (nonidentical or ‘fraternal’) twins. This would also account for observations that in genetically identical inbred mouse lines, animals derived from a single fertilized egg (a model for human MZ twins) are more similar than normal sib-pairs, despite both pair types sharing identical DNA sequences and environments (Petronis, 2006).

In a recent review of twin studies in hypoxia (MacLeod et al., 2013), a number of examples of discordant twin-pairs are discussed. These could be due to epigenetic modulation of molecular or physiological responses to environmental hypoxia (reviewed in Watson et al., 2010), possibly working through the HIF pathway through the mechanisms discussed earlier. Comparing hypoxia responses between twin types could provide insights into role of epigenetics in altitude acclimatization (or adaptation) as well as identify specific genes and variations that contribute to hypoxia sensitivity (i.e., susceptibility to altitude illness) or tolerance.

Future Directions: Epigenetics and Altitude Adaptation

Whether epigenetic changes play a role in altitude acclimatization or adaptation is as yet unknown. The ability of a parent living in a hypoxic environment to improve their unborn child's (or grandchild's) ability to uptake, transport and utilize oxygen would have greatly facilitated our species' occupation of the highlands. One mechanism by which this could have occurred is through modifications of the placental interface. Placental function altered in response to altitude may be a key adaptive strategy (Moore et al., 2004). One of the functional categories of genes differentially expressed in the placenta (at 15.5 days of gestation) of mice exposed to 10.5% O₂ for 48 hours were genes involved in DNA methylation (Gheorghe et al., 2007), suggesting that embryonic exposure to hypoxia may induce additional epigenetic changes. Many imprinted genes are expressed in the mouse during development, both embryonically and in the placenta, where they appear to regulate nutrient demand and supply (Cetin et al., 2004). Whether the placental delivery of oxygen is modulated by epigenetics is unknown; however, imprinted genes might be particularly susceptible to epigenetic changes as they (like X-linked genes) are normally monoallelically expressed.

Gestational hypertension and preeclampsia may contribute to lower birth weights in the highlands in both the developed (e.g. Colorado, 2744 – 3350 m; (Jensen and Moore 1997) and developing world (e.g., LaPaz, Bolivia 3600 m; Keyes et al., 2003). However, babies born to highland native mothers tend to be bigger than those born to non-natives (Bennett et al., 2008). Placental function seems to be affected by epigenetically marked genes, but whether this confers a superior uterine environment in highland women has yet to be investigated.

Determining whether epigenetic changes have contributed significantly to human acclimatization or adaptations to hypoxia will be challenging. One potential confounding factor will be the effects of other environmental changes. In addition to the well-studied effects of nutrition, many of the changes associated with population density (e.g., infectious pathogens), urbanization (e.g., pollutants, particulate matter, allergens), adoption of a ‘technological’ society (metals, plastics (BPAs)), and cultural changes (tobacco use) can affect epigenetic patterns (reviewed in Ho et al., 2012). Furthermore, as epigenetic marks can be cell-type specific, differences in sample composition can confound analysis, especially when working with tissues that have shifting cell populations, such as blood or placenta.

Conclusions

Epigenetic programming allows adaptation to changing environments via mechanisms that allow the adaptive responses to be maintained in future cells and potentially future generations. As epigenetics allows cells to modify their expression patterns, and then pass on this response to their descendants, any adaptive consequences could manifest in an individual long after exposure or even be passed on to their offspring. From an altitude acclimatization perspective, this could facilitate responses during subsequent ascents; whereas, from an altitude adaptation perspective, epigenetics may allow “preprogramming” of offspring to more efficiently deal with a hypoxic environment previously experienced by a parent.

Substantial resources have been invested in understanding the human epigenome, including genome-wide scans for epigenetic markers (e.g., Brinkman et al.) and much data is available to researchers though sites such as ENCODE (http://www.genome.gov/Encode/). Unlike DNA sequence variation, which can be studied in any cell, epigenetic marks may be cell or tissue specific. While this presents a challenge in human studies (as sampling is usually limited to peripheral cells such as blood), advances have been made towards understanding epigenetic responses to environmental stressors such as caloric restriction and low socioeconomic status (Lam et al., 2012). Whether an individual's, or a population's, epigenome contributes to the capacity to tolerate hypoxia has yet to be systematically investigated; however, given the potential of epigenetics to revolutionize our understanding of adaptive biology, this is a challenge that needs be accepted.